Hi Everyone,
I am always grateful for the help that is shared. I have two questions.
The Left channel is mistreating the input waveform as it passes through the circuit. Last night, I signal traced a clean square wave up until the 3 band EQ portion of the circuit. The output signal looks more like a triangle wave, with sloped rise and fall times - see photo. The EQ uses opamps. Which part of the circuit should I check for faults?
My second question, why does the schematic only show the Left channel of a stereo input? See schematic, attached. Thanks!
I am always grateful for the help that is shared. I have two questions.
The Left channel is mistreating the input waveform as it passes through the circuit. Last night, I signal traced a clean square wave up until the 3 band EQ portion of the circuit. The output signal looks more like a triangle wave, with sloped rise and fall times - see photo. The EQ uses opamps. Which part of the circuit should I check for faults?
My second question, why does the schematic only show the Left channel of a stereo input? See schematic, attached. Thanks!
Attachments
You realise a band limited square wave will not look like a square wave ?
You don't say:
a) How the EQ was set during the test. (Flat ?)
b) what the centre frequencies of each band of the EQ are,
c) what the fundamental frequency of your square wave is.
Without knowing those it's very possible that the amplifier is working absolutely normally.
For example if you put a 5Khz square wave in the harmonics will extend to hundreds of Khz - they may pass through the initial pre-amp stages but be band limited by the graphic EQ to 20Khz or so and this will distort a square wave into something that looks more like a triangle wave.
As tests go, passing a square wave of indeterminate frequency through an amplifier is not a particularly useful one.
You don't say:
a) How the EQ was set during the test. (Flat ?)
b) what the centre frequencies of each band of the EQ are,
c) what the fundamental frequency of your square wave is.
Without knowing those it's very possible that the amplifier is working absolutely normally.
For example if you put a 5Khz square wave in the harmonics will extend to hundreds of Khz - they may pass through the initial pre-amp stages but be band limited by the graphic EQ to 20Khz or so and this will distort a square wave into something that looks more like a triangle wave.
As tests go, passing a square wave of indeterminate frequency through an amplifier is not a particularly useful one.
Yes, I took all of that for granted.
1kHz square wave
150mVrms @ CD input, recommended by the schematic
EQ set at center point (no gain or cut at selected frequencies)
Most importantly, the preamp has a "source direct" switch, which shows the 1kHz as a near perfect square. Each test that was performed did compare the Equalized version of the signal to the "source direct" version of the signal. On the Right channel, the Equalized version looked much closer to an actual square wave.
^^Both the left and right channels perform fine when "source direct" is selected. But, the two channels are very different when EQ is engaged. That is why I asked the second question as well.
1kHz square wave
150mVrms @ CD input, recommended by the schematic
EQ set at center point (no gain or cut at selected frequencies)
Most importantly, the preamp has a "source direct" switch, which shows the 1kHz as a near perfect square. Each test that was performed did compare the Equalized version of the signal to the "source direct" version of the signal. On the Right channel, the Equalized version looked much closer to an actual square wave.
^^Both the left and right channels perform fine when "source direct" is selected. But, the two channels are very different when EQ is engaged. That is why I asked the second question as well.
Can someone correct me if I am wrong? Here are my assumptions about the problem.
The issue is a matter of phase shift or a change in RC combinations within the EQ or NFB circuits. If I have zero change in gain, and the power supply voltages are all good, than I think the opamps are working correctly.
As for solutions, I can inject a 1kHz sq wave signal into the amp AFTER the affected portion of the circuit - this will narrow down the suspected area so that I can troubleshoot the components individually with a DMM and an LCR meter.
The issue is a matter of phase shift or a change in RC combinations within the EQ or NFB circuits. If I have zero change in gain, and the power supply voltages are all good, than I think the opamps are working correctly.
As for solutions, I can inject a 1kHz sq wave signal into the amp AFTER the affected portion of the circuit - this will narrow down the suspected area so that I can troubleshoot the components individually with a DMM and an LCR meter.
It looks like treble roll-off, or at least some emphasis of low frequencies overall, which means it could also be bass boost.
Switch between bypass and tone controls in, and use sine waves to make sure which it is. At that point, you can just concentrate on the components associated with the named control.
Why only one channel shown? It's a common short-hand, as both channels are the same...still, you'd think that they'd specifically list the equivalent components for the other channel...sometimes it isn't necessary if it's on an equivalent PCB with the same numbering.
Switch between bypass and tone controls in, and use sine waves to make sure which it is. At that point, you can just concentrate on the components associated with the named control.
Why only one channel shown? It's a common short-hand, as both channels are the same...still, you'd think that they'd specifically list the equivalent components for the other channel...sometimes it isn't necessary if it's on an equivalent PCB with the same numbering.
UPDATE: Square wave now passes the test on both channels, after 3 capacitors (C301, C302, C304) were changed in the "simulated stereo" op amp circuit. One of these three brought the rising treble edge back to where it is supposed to be (see photo #2). A fourth capacitor (C303 now controls the bass frequency. This is really interesting, can someone explain what is happening?
Schematic and two photos attached. Photo #1 is of a 10nF (stock value) in the C303 position, and photo #2 a 1nF in the C303 position, respectively.
*If a 10nF, as the design has intended, no longer suits the needs of the circuit (see photo #1, with 10nF installed), than what component value has gone "bad"? Is the op amp bad? The only thing I can see that is still original in this circuit are the resistors and the op amp.
**If a 1nF does a nice job in C303 by bringing the bass back, should I try a 100pF in the C303 spot to get rid of the treble spike? As you can imagine, I don't understand how C303 functions in the circuit, and I'm still not sure if something else is wrong. I am sure that the amp is telling me that a 1nF in C303 is moving in the right direction, but I also know that I should not attempt to re-engineer what I do not comprehend.
Schematic and two photos attached. Photo #1 is of a 10nF (stock value) in the C303 position, and photo #2 a 1nF in the C303 position, respectively.
*If a 10nF, as the design has intended, no longer suits the needs of the circuit (see photo #1, with 10nF installed), than what component value has gone "bad"? Is the op amp bad? The only thing I can see that is still original in this circuit are the resistors and the op amp.
**If a 1nF does a nice job in C303 by bringing the bass back, should I try a 100pF in the C303 spot to get rid of the treble spike? As you can imagine, I don't understand how C303 functions in the circuit, and I'm still not sure if something else is wrong. I am sure that the amp is telling me that a 1nF in C303 is moving in the right direction, but I also know that I should not attempt to re-engineer what I do not comprehend.
Bump,
I will test a new value of capacitor in the C303 position tonight. I would like to receive some feedback from someone with more experience and wisdom, so that I can learn whether this is an "okay" thing to do.
I have read about amplifier repairmen who, at times, must add a small amount of capacitance to a solid state device to prevent it from oscillating. I understand that this is because each component varies in its exact value of L, C, and R. What I am noticing in this Onkyo is different than that, as the amplifier Q301 is not oscillating, but does appear to be reflecting the value of the RC filters in place. I am wondering if an op amp can change its input C or input Z, or if an op amp can become leaky.
I will test a new value of capacitor in the C303 position tonight. I would like to receive some feedback from someone with more experience and wisdom, so that I can learn whether this is an "okay" thing to do.
I have read about amplifier repairmen who, at times, must add a small amount of capacitance to a solid state device to prevent it from oscillating. I understand that this is because each component varies in its exact value of L, C, and R. What I am noticing in this Onkyo is different than that, as the amplifier Q301 is not oscillating, but does appear to be reflecting the value of the RC filters in place. I am wondering if an op amp can change its input C or input Z, or if an op amp can become leaky.
Simulated Stereo Circuit Tests:
1. Swap Op with another (same type) found in balance control circuit. No effect.
2. Installed a 470pF in C303. "Squeezed" all of the rising square wave energy into a narrower and taller spike, bass was the same as 1nF. *SPIKE IS REALLY A VALLEY? [edit, photo of results when 470pF was in circuit is not available]
3. Measured all resistors, they are 1% or better.
4. Scope'd every in and out on the op amp and related circuit.
5. Supply voltages are +/-12V. Schem asks for 12.5V, and it is a regulated supply.
6. Scope'd Supply rails, they are clean.
7. Compared signals into inv/non-inv (pins 3 + 4). When they look the same going in, they come out with spike, after the summing process.
8. Measured V across R304/R305 Op amp input resistors - R304 = 100uA, R305 = 50uA.
1. Swap Op with another (same type) found in balance control circuit. No effect.
2. Installed a 470pF in C303. "Squeezed" all of the rising square wave energy into a narrower and taller spike, bass was the same as 1nF. *SPIKE IS REALLY A VALLEY? [edit, photo of results when 470pF was in circuit is not available]
3. Measured all resistors, they are 1% or better.
4. Scope'd every in and out on the op amp and related circuit.
5. Supply voltages are +/-12V. Schem asks for 12.5V, and it is a regulated supply.
6. Scope'd Supply rails, they are clean.
7. Compared signals into inv/non-inv (pins 3 + 4). When they look the same going in, they come out with spike, after the summing process.
8. Measured V across R304/R305 Op amp input resistors - R304 = 100uA, R305 = 50uA.
Last edited:
- Status
- Not open for further replies.